1 / 73

CHAPTER 1

CHAPTER 1. Position and Orientation Transformations and Robot Kinematics of Position. After this chapter, the students are expected to learn the following:. Definitions of positions and orientations of rigid bodies Analysis of different representations for orientation

gaius
Download Presentation

CHAPTER 1

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. CHAPTER 1 Position and Orientation Transformations and Robot Kinematics of Position

  2. After this chapter, the students are expected to learn the following: • Definitions of positions and orientations of rigid bodies • Analysis of different representations for orientation • Transformation of coordinates

  3. After this chapter, the students are expected to learn the following: • Determination of new positions and orientations after a sequence of rigid body motions • Kinematics modeling of robotic manipulators • Forward and inverse kinematics of position

  4. SPATIAL DESCRIPTION • used to specify spatial attributes of various objects with which a manipulation system deals • universe coordinate frame is implicit • use Cartesian coordinate frames

  5. zA o yA APoz APox APoy xA POSITION • attribute of a point (APox , APoy, APoz) = Cartesian coordinates of pt.0 expressed in frame A Position vector - (coords has magnitude & direction, origin impt)

  6. ORIENTATION • attribute of a body • attach a Cartesian coordinate frame to the body • orientation of each axis of the Cartesian coordinate frame • free vectors = orientation vectors = describe orientation • magnitude & origin not important • only direction is important

  7. zA AxB AyB AzB yA xA ORIENTATION • (AxB, AyB, AzB) • unit vectors along each • axis of frame B • free vectors, only direction • is relevant

  8. Rotation Matrix • Since each column of the rotation matrix represents unit vectors along the x, y and z directions of the Cartesian coordinates, then they should be orthogonal to each other • Each column obeys unit length constraints • Cartesian coord frame right hand rule ARB is a Proper Orthogonal Matrix ARB-1 = ARBT = BRA or ARBARBT = I

  9. Rotation Matrix • All possible orientations of a rigid body (i.e. coordinate frame attached to the body) can be uniquely specified by a rotation matrix • 3 independent parameters are only needed

  10. (1,0,0) in Frame B zA o Same point has coordinates (AxB,x , AxB,y , AxB,z) in Frame A xB yB xA zB Coordinate Transformations & Rigid Body Motion • Similarly, (0,1,0) in Frame B has • coordinates • 2nd column of ARB in Frame A • And (0,0,1) in Frame B has • coordinates • 3rd column of ARB in Frame A yA

  11. Coordinate Transformations & Rigid Body Motion Generalizing – a point, O, with coordinates in Frame B can be transformed to coordinates in Frame A if the relative orientation between A & B (ARB) is known APo = ARB BPo Coordinates of O in B Coordinates of O in A Note the ASSUMPTION : – origins of Frames A & B are coincident – only difference is in orientation

  12. zA’ o yB z xB yA’ y A x zB xA’ Coordinate Transformations & Rigid Body Motion What if the two origins are not coincident? – i.e. , translation between frames A & B A’ = frame parallel to A (only different origins) A’Po = A’RBBPo (from before)

  13. Coordinate Transformations & Rigid Body Motion APo = APB + A’Po can only add two position vectors if they are expressed in Frames that are parallel to each other A is parallel to A’ – so that adding respective coordinates make sense Therefore APo = APB + ARBBPo since ARB = A’RB because A is // A’ Transforming O to a frame // to A Then adding the relative displacement of A & B

  14. zA xB yB o yA zB o yB xA zB xB Duality (of Coordinate Transformation) with Rigid Body Motion • Imagine a Rigid Body (Pyramid) • whose apex is at pt. O • Frame B is attached to the • rigid body

  15. zA xB1 xBo yBo o o1 yA zB1 yB1 xA zBo Duality (of Coordinate Transformation) with Rigid Body Motion • The body B is initially at ARBo and APo = ARBoBPo • Body B undergoes a motion such that the new Position & • Orientation of B are ARB1 & APB1 • New Coordinates of O in A, APO1 can be found : • APo1 = APB1 + ARB1B1 Po

  16. zAo= A zA1 yA1  yAo= A xAo= A xA1 Elementary (Basic, Fundamental) Motions • Initially, Body is at Ao • A undergoes a Rotation about xAo • by  & A is now at A1 after the • rotation Similarly, a Rotation about y by  :

  17. zA1 zAo yA1 AoPA1 yAo xA1 xAo Elementary (Basic, Fundamental) Motions Similarly, a rotation about z by  : Translation – Body A undergoes pure translation (only) Coordinates of A1 in Ao

  18. Homogeneous Transformations • created so that rotations and translations are treated uniformly. (i.e. as matrix multiplications) • vectors have 4 components • scaling factor = 0 (free vectors, unit vectors along axis) • scaling factor = 1 (position vectors)

  19. Homogeneous Transformations APC = ARBBPC + APB = homogeneous transformation matrix = describes the position and orientation of frame B in frame A Chain rule : ATC = ATBBTC

  20. ARB X = I X = ARB-1 = ARBT = BRA ARB Y + APB = 0 ARB Y = -APB Y = -ARB-1APB = -ARBTAPB Inverse of a Homogeneous Transformation Matrix Given ATB, Find BTA ATB BTA = I

  21. Inverse of a Homogeneous Transformation Matrix Note that ( n0a )T = ARBT ARBT APB = n• APB 0 • APB a • APB

  22. zA zB cube of unit length c xB yA xA Example : Given : Initial Position & Orientation of cube specified by ATB Find : new coordinates of pt. c (corner of cube) after the cube is rotated by  about zA

  23. APB’ B’PC = B’’PC since C is attached rigidly to B Example :

  24. Inverse Transf. : Given ARB, Find , ,  Other Orientation Representation Euler angles , , 

  25. Mathematical Singularity occurs when az = 1, • and  describe the same rotation & cannot be computed separately Other Orientation Representation

  26. y x ARB = Rot ( z,  ) Rot ( y,  ) Rot ( x,  ) z Roll Pitch Yaw ROLL - PITCH - YAW Angles( , ,  )

  27. Mathematical Singularity occurs when nz = 1 • &  describe the same rotation & cannot be computed separately Inverse Transformation

  28. Any rotation matrix can be represented by rotation  of space about a fixed axis k = zA  k A B after rotation K is a vector expressed in A yA xA Four-Parameter Representation For Orientation • quadruple of ordered real parameters consisting of one • scalar (angle of rotation) and one vector (axis of rotation) ARB = Rot ( k ,  )

  29. where vers  = 1 – cos  Try proving this Hint : imagine to be the x, y or z axis of a new frame. Rot ( h , ) then becomes one of the elementary rotations [ Rot ( x,  ) , Rot ( y,  ) or Rot ( z,  ) ] Four-Parameter Representation For Orientation

  30. Given ARB Find k,  previous page given From the diagonal elements, Trace (ARB) = 1 + 2cos cos = ½( Trace (ARB) – 1 ) Two solutions for   = A +A -A Inverse Problem

  31. From : nz & ax elements From : oz & ay elements Valid only for sin  0 From : ny & ox elements Inverse Problem

  32. If  = 0º Identity k = any axis (infinite sol’ns) If  = 180º = symmetric = Tr (ARB) = -3 + 2 = -1 = 1 + 2cos Inverse Problem

  33. If no off diagonal term is zero, 2 solution for k If one solution is , the other is Inverse Problem If all off diagonal terms are , one of kx, ky or kz = 1 & the rest of the components of k are . The component of k that is 1 has a value of 1 in the diagonal of ARB.

  34. 0 = cos /2 Note : 02 + 12 + 22 + 32 = 1 Euler Parameters, or Quaternions

  35. ( 2 Sol’ns ) (  = 180º ) Inverse Transformation Given ARB, Find 0, 1, 2, 3 Trace (ARB) = 2(12 + 22 + 32) – 3 = 2 – 3 = -1

  36. 2 solutions If one sol’n is 1, 2, 3 other sol’n is -0, -2, -3 Inverse Transformation All diagonal terms are either : – all zero, or – all non-zero

  37. Robot Kinematic Modeling THE DENAVIT-HARTENBERG REPRESENTATION In the robotics literature, the Denavit-Hartenberg (D-H) Representation has been used, almost universally, to derive the kinematic description of robotic manipulators. The appeal of the D-H representation lies in its algorithmic approach. In this handout, we provide an algorithm for the assignment of robotic coordinate frames, highlight the conventions associated with the D-H approach, and exemplify the development through the Puma and Stanford manipulators.

  38. Robot Kinematic Modeling STEP 1 : Number the Robot Joints and Links Robotic manipulators are articulated, open kinematic chains of N rigid bodies (links) which are connected serially by joints. The links are numbered consecutively from the base (link 0) to the final end (link N). The joints are the points of articulation between the links and are numbered from 1 to N so that joint i connects links (i-1) and i. Each joint provides one degree-of-freedom which can either be a rotation or translation.There is no joint at the end of the final link.

  39. Robot Kinematic Modeling • STEP 2 : Assign Link Coordinate Frames • To describe the geometry of robot motion, we assign a Cartesian coordinate frame (Oi;xi,yi,zi) to each link, as follows: • the zi axis is directed along the axis of motion of • joint (i + 1), that is, link (i + 1) rotates about or • translates along zi; • the xi axis lies along the common normal from • the zi-1 axis to the zi axis (if zi-1 is parallel to zi, • then xi is specified arbitrarily, subject only to xi • being perpendicular to zi); and

  40. Robot Kinematic Modeling • STEP 2 : Assign Link Coordinate Frames • the yi axis completes the right-handed coordinate • system. • The origin of the robot base frame O0 can be placed anywhere in the supporting base and the origin of the last (end-effector) coordinate frame ON is specified by the geometry of the end-effector.

  41. Robot Kinematic Modeling STEP 3 : Define the Joint Coordinates The joint coordinate qi is the angular displacement around zi-1if joint i is rotational, or the linear displacement along zi-1 if joint i is translational. The N-dimensional space defined by the joint coordinates (q1,…,qN) is called the configuration space of the N DOF mechanism.

  42. Robot Kinematic Modeling • STEP 4 : Identify the Link Kinematic Parameter • In general, four elementary transformations are required to relate the i-th coordinate frame to the • (i-1)-th coordinate frame: • Rotate an angle of i (in the right-handed sense) • about the zi-1 axis, so that the xi-1 axis is parallel • to the xi axis. • Translate a distance of ri along the positive • direction of the zi-1 axis, to align the xi-1 axis with • the xi axis.

  43. Robot Kinematic Modeling • STEP 4 : Identify the Link Kinematic Parameter • Translate a distance of di along the positive • direction of the xi-1 = xi axis, to coalesce the • origins Oi-1 and Oi. • Rotate an angle of i (in the right-handed sense) • about the xi-1 = xi axis, to coalesce the two • coordinate systems. • The i-th coordinate frame is therefore characterized by the four D-H kinematic link parameters i, ri, di and i. If joint i is rotational, then qi = i, and i, di and ri are constant parameters which depend upon the

  44. Robot Kinematic Modeling STEP 4 : Identify the Link Kinematic Parameter geometric properties and configuration of link i. If joint i is translational, then qi = ri, and di, i and i are constant parameters which depend upon the configuration of link i. For both rotational and translational joints, ri and i are the distance and angle between links (i – 1) and i; di and i are the length and twist of link i.

  45. Robot Kinematic Modeling STEP 4 : Identify the Link Kinematic Parameter

  46. Robot Kinematic Modeling STEP 5 : Define the Link Transformation Matrices The position and orientation of the i-th coordinate frame can be expressed in the (i – 1)-th coordinate frame by the following homogeneous transformation matrix: Ai = Rot(z, ) Trans(0, 0, ri) Trans(di, 0, 0) Rot(x, )

  47. STEP 6 : Compute the Forward Transformation Matrix The position and orientation of the end-effector coordinate frame is expressed in the base coordinate frame by the forward transformation matrix: 0TN(q1, q2,…, qN) = 0TN = A1A2…AN = Robot Kinematic Modeling

  48. Robot Kinematic Modeling EXAMPLE 1: The Puma Robot

  49. Robot Kinematic Modeling EXAMPLE 2: The Stanford Arm

  50. q3 q2 q1 z y x Forward Kinematic Problem Given: q1, q2, q3…. (joint positions) Find: End-Effector position PE and orientation RE

More Related